{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import scipy.stats as sp\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import warnings\n",
    "import chardet\n",
    "from collections import Counter\n",
    "import dabest\n",
    "from statsmodels.stats.proportion import proportions_ztest\n",
    "import statsmodels.api as sm\n",
    "warnings.filterwarnings('ignore')\n",
    "sns.set_style(\"white\")\n",
    "sns.set_context(\"paper\")\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "slideshow": {
     "slide_type": "notes"
    }
   },
   "outputs": [],
   "source": [
    "#with open('cleaned_data.csv', 'rb') as f:\n",
    "#    coding = chardet.detect(f.read())  # or readline if the file is large\n",
    "df = pd.read_csv('./data.csv',sep=',',index_col='ResponseId')\n",
    "order = ['T1','T2','T3','T4']\n",
    "T1_mask = df.treatment=='T1'\n",
    "T2_mask = df.treatment=='T2'\n",
    "T3_mask = df.treatment=='T3'\n",
    "T4_mask = df.treatment=='T4'\n",
    "T_dict = {0:'T1',1:'T2',2:'T3',3:'T4'}\n",
    "df['donation_percentage'] = df.contribution*100"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data description"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number of participants 2500\n",
      "number of invalid participants is 163 which is: 6.52 %\n",
      "we will use:  2337\n"
     ]
    }
   ],
   "source": [
    "print('number of participants',len(df))\n",
    "print('number of invalid participants is',len(df[df.valid==False]), 'which is:',len(df[df.valid==False])/len(df)*100.,'%')\n",
    "df = df[df.valid==True]\n",
    "print('we will use: ',len(df))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, '# Participants')"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(8, 6))\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "sns.countplot(x=\"gender\", data=df,ax=ax,order=['female','male','other','not reported'])\n",
    "for i,p in enumerate(ax.patches):\n",
    "    total = len(df)\n",
    "    height = p.get_height()\n",
    "    ax.text(p.get_x()+p.get_width()/2.,\n",
    "            height + 10,\n",
    "            '{:1.2f}%'.format(height/total*100),\n",
    "            ha=\"center\",fontsize=15) \n",
    "ax.set_ylim(0,1300)    \n",
    "ax.tick_params(labelsize=15)\n",
    "ax.set_xlabel('Gender',size=19)\n",
    "ax.set_ylabel('# Participants',size=19)\n",
    "#plt.savefig('./gender_breakdown.pdf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "average age 36.24207369323051\n",
      "median age 34.0\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, '# Participants')"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print('average age',df.age.mean())\n",
    "print('median age',df.age.median())\n",
    "\n",
    "fig = plt.figure(figsize=(8, 6))\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "ax.axvline(df.age.mean())\n",
    "#ax.axvline(df.age.median(),c='r')\n",
    "\n",
    "sns.distplot(df.age[~np.isnan(df.age)],ax=ax,kde=False,rug=False)\n",
    "#for i,p in enumerate(ax.patches):\n",
    "#    total = len(df)\n",
    "#    height = p.get_height()\n",
    "#    ax.text(p.get_x()+p.get_width()/2.,\n",
    "#            height + 10,\n",
    "#            '{:1.2f}%'.format(height/total*100),\n",
    "#            ha=\"center\") \n",
    "#ax.set_ylim(0,200)    \n",
    "ax.tick_params(labelsize=15)\n",
    "ax.set_xlabel('Age',size=19)\n",
    "ax.set_ylabel('# Participants',size=19)\n",
    "#plt.savefig('./age_distribution.pdf')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Treatment level Analysis\n",
    "<p>Error bars indicate standard error of the mean</p>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "custom_palette = [\"windows blue\", \"amber\", \"greyish\", \"black\"] #\"green blue\", \"dull green\", \"faded green\",  \n",
    "sns.set_palette(sns.xkcd_palette(custom_palette))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(10, 6))\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "\n",
    "g = sns.barplot(x=\"treatment\", y=\"donation_percentage\", data=df, ax=ax,order=order,ci=68, errwidth=1, capsize=0.1) \n",
    "\n",
    "ax.tick_params(labelsize=15)\n",
    "ax.set_xlabel('Condition',size=19)\n",
    "ax.set_ylabel('Avg. donation',size=19)\n",
    "ax.set_ylim(df.contribution.min(),18)\n",
    "plt.savefig('./avg_donation.pdf')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<strong>Significance:</strong> using Mann–Whitney U test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "average\n",
      "T1 0.15592274678111584\n",
      "T2 0.13572230014025236\n",
      "T3 0.11582743988684586\n",
      "T4 0.013394495412844038\n",
      "Median\n",
      "T1 0.0\n",
      "T2 0.0\n",
      "T3 0.0\n",
      "T4 0.0\n",
      "signiciance\n",
      "T1 > T2 p-val: MannwhitneyuResult(statistic=235887.0, pvalue=0.027057870505368765)\n",
      "T1 > T3 p-val: MannwhitneyuResult(statistic=220531.5, pvalue=4.1332729554323674e-05)\n",
      "T2 > T3 p-val: MannwhitneyuResult(statistic=238020.0, pvalue=0.019076863058705568)\n",
      "T3 > T4 p-val: MannwhitneyuResult(statistic=55559.0, pvalue=6.0497003101169214e-15)\n",
      "% improvement\n",
      "T1 vs T2 0.14883660695396997\n",
      "T1 vs T3 0.3461641467120389\n",
      "T2 vs T3 0.17176292830819861\n"
     ]
    }
   ],
   "source": [
    "print('average')\n",
    "print('T1', df[T1_mask]['contribution'].mean())\n",
    "print('T2', df[T2_mask]['contribution'].mean())\n",
    "print('T3', df[T3_mask]['contribution'].mean())\n",
    "print('T4', df[T4_mask]['contribution'].mean())\n",
    "\n",
    "print('Median')\n",
    "print('T1', df[T1_mask]['contribution'].median())\n",
    "print('T2', df[T2_mask]['contribution'].median())\n",
    "print('T3', df[T3_mask]['contribution'].median())\n",
    "print('T4', df[T4_mask]['contribution'].median())\n",
    "\n",
    "print('signiciance')\n",
    "print('T1 > T2 p-val:',sp.mannwhitneyu(df[T1_mask]['contribution'],df[T2_mask]['contribution']))\n",
    "print('T1 > T3 p-val:',sp.mannwhitneyu(df[T1_mask]['contribution'],df[T3_mask]['contribution']))\n",
    "print('T2 > T3 p-val:',sp.mannwhitneyu(df[T2_mask]['contribution'],df[T3_mask]['contribution']))\n",
    "print('T3 > T4 p-val:',sp.mannwhitneyu(df[T3_mask]['contribution'],df[T4_mask]['contribution']))\n",
    "\n",
    "print('% improvement')\n",
    "print('T1 vs T2', (df[T1_mask]['contribution'].mean()-df[T2_mask]['contribution'].mean())/df[T2_mask]['contribution'].mean())\n",
    "print('T1 vs T3', (df[T1_mask]['contribution'].mean()-df[T3_mask]['contribution'].mean())/df[T3_mask]['contribution'].mean())\n",
    "print('T2 vs T3', (df[T2_mask]['contribution'].mean()-df[T3_mask]['contribution'].mean())/df[T3_mask]['contribution'].mean())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<strong>Giving %:</strong> Participants that gave 'something'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['gave_something'] = df.contribution > 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Between T1 and T2 (2.112495168747219, 0.01732199995708073)\n",
      "Between T1 and T3 (5.079495666159294, 1.8921907865477295e-07)\n",
      "Between T2 and T3 (3.0282959455776948, 0.0012296855360988002)\n",
      "Between T3 and T4 (19.910284429122683, 1.6571402165662913e-88)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(5, 6))\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "sns.countplot(x=\"treatment\", data=df[df.gave_something==True],ax=ax,order=order)\n",
    "for i,p in enumerate(ax.patches):\n",
    "    total = sum((df.treatment==T_dict[i]))\n",
    "    height = p.get_height()\n",
    "    ax.text(p.get_x()+p.get_width()/2.,\n",
    "            height + 4,\n",
    "            '{:1.2f}%'.format(height/total*100),\n",
    "            ha=\"center\", fontsize=15) \n",
    "ax.set_ylim(0,370)    \n",
    "ax.tick_params(labelsize=15)\n",
    "ax.set_xlabel('Condition',size=19)\n",
    "ax.set_ylabel('# Participants donated > 0',size=19)\n",
    "\n",
    "##statistical test\n",
    "n_gave_T1 = sum(df[df.treatment=='T1'].gave_something)\n",
    "n_NOTgave_T1 = len(df[df.treatment=='T1']) - n_gave_T1\n",
    "n_gave_T2 = sum(df[df.treatment=='T2'].gave_something)\n",
    "n_NOTgave_T2 = len(df[df.treatment=='T2']) - n_gave_T2\n",
    "n_gave_T3 = sum(df[df.treatment=='T3'].gave_something)\n",
    "n_NOTgave_T3 = len(df[df.treatment=='T3']) - n_gave_T3\n",
    "n_gave_T4 = sum(df[df.treatment=='T4'].contribution == 0.5)\n",
    "n_NOTgave_T4 = len(df[df.treatment=='T4']) - n_gave_T4\n",
    "\n",
    "\n",
    "print('Between T1 and T2',proportions_ztest(n_gave_T1,n_NOTgave_T1+n_gave_T1,\n",
    "                                            value=n_gave_T2/(n_gave_T2+n_NOTgave_T2),\n",
    "                                            alternative='larger'))\n",
    "\n",
    "print('Between T1 and T3',proportions_ztest(n_gave_T1,n_NOTgave_T1+n_gave_T1,\n",
    "                                            value=n_gave_T3/(n_gave_T3+n_NOTgave_T3),\n",
    "                                            alternative='larger'))\n",
    "\n",
    "\n",
    "print('Between T2 and T3',proportions_ztest(n_gave_T2,n_NOTgave_T2+n_gave_T2,\n",
    "                                            value=n_gave_T3/(n_gave_T3+n_NOTgave_T3),\n",
    "                                            alternative='larger'))\n",
    "print('Between T3 and T4',proportions_ztest(n_gave_T3,n_NOTgave_T3+n_gave_T3,\n",
    "                                            value=n_gave_T4/(n_gave_T4+n_NOTgave_T4),\n",
    "                                            alternative='larger'))\n",
    "\n",
    "plt.savefig('./gave_something.pdf')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<strong>Giving %:</strong> Participants that gave 50/50"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Between T1 and T2 (3.840147386833523, 6.148023366898333e-05)\n",
      "Between T1 and T3 (4.775542323127287, 8.961179548466965e-07)\n",
      "Between T2 and T3 (1.0781888694936892, 0.14047473905634916)\n",
      "Between T3 and T4 (10.412221455004632, 1.0903627737918944e-25)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0, 370)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(5, 6))\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "sns.countplot(x=\"treatment\", data=df[df.contribution == 0.5],ax=ax,order=order)\n",
    "for i,p in enumerate(ax.patches):\n",
    "    if i == 3:\n",
    "        height =0.1\n",
    "        ax.text(p.get_x()+p.get_width()/2.,\n",
    "                height + 4,\n",
    "                '{:1.2f}%'.format(0),\n",
    "                ha=\"center\",fontsize=15)\n",
    "    else:\n",
    "        total = sum((df.treatment==T_dict[i]))\n",
    "        height = p.get_height()\n",
    "        ax.text(p.get_x()+p.get_width()/2.,\n",
    "                height + 3,\n",
    "                '{:1.2f}%'.format(height/total*100),\n",
    "                ha=\"center\",fontsize=15) \n",
    "        \n",
    "        \n",
    "n_gave_T1 = sum(df[df.treatment=='T1'].contribution == 0.5)\n",
    "n_NOTgave_T1 = len(df[df.treatment=='T1']) - n_gave_T1\n",
    "n_gave_T2 = sum(df[df.treatment=='T2'].contribution == 0.5)\n",
    "n_NOTgave_T2 = len(df[df.treatment=='T2']) - n_gave_T2\n",
    "n_gave_T3 = sum(df[df.treatment=='T3'].contribution == 0.5)\n",
    "n_NOTgave_T3 = len(df[df.treatment=='T3']) - n_gave_T3\n",
    "n_gave_T4 = sum(df[df.treatment=='T4'].contribution == 0.5)\n",
    "n_NOTgave_T4 = len(df[df.treatment=='T4']) - n_gave_T4\n",
    "\n",
    "\n",
    "print('Between T1 and T2',proportions_ztest(n_gave_T1,n_NOTgave_T1+n_gave_T1,\n",
    "                                            value=n_gave_T2/(n_gave_T2+n_NOTgave_T2),\n",
    "                                            alternative='larger'))\n",
    "\n",
    "print('Between T1 and T3',proportions_ztest(n_gave_T1,n_NOTgave_T1+n_gave_T1,\n",
    "                                            value=n_gave_T3/(n_gave_T3+n_NOTgave_T3),\n",
    "                                            alternative='larger'))\n",
    "\n",
    "\n",
    "print('Between T2 and T3',proportions_ztest(n_gave_T2,n_NOTgave_T2+n_gave_T2,\n",
    "                                            value=n_gave_T3/(n_gave_T3+n_NOTgave_T3),\n",
    "                                            alternative='larger'))\n",
    "print('Between T3 and T4',proportions_ztest(n_gave_T3,n_NOTgave_T3+n_gave_T3,\n",
    "                                            value=n_gave_T4/(n_gave_T4+n_NOTgave_T4),\n",
    "                                            alternative='larger'))\n",
    "\n",
    "    \n",
    "ax.tick_params(labelsize=15)\n",
    "ax.set_xlabel('Condition',size=19)\n",
    "ax.set_ylabel('# Participants donated %50',size=19)\n",
    "ax.set_ylim(0,370)\n",
    "#plt.savefig('./gave50_percent.pdf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Between T1 and T2 (-3.0359987089951486, 0.9988012980457135)\n",
      "Between T1 and T3 (-1.3058362738420581, 0.9041958792149931)\n",
      "Between T2 and T3 (1.1447759976007794, 0.1261509802889867)\n",
      "Between T3 and T4 (2.8445666401983494, 0.0022235941049615483)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0, 370)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(5, 6))\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "sns.countplot(x=\"treatment\", data=df[df.contribution > 0.5],ax=ax,order=order)\n",
    "for i,p in enumerate(ax.patches):\n",
    "    if i == 3:\n",
    "        height =0.1\n",
    "        ax.text(p.get_x()+p.get_width()/2.,\n",
    "                height + 4,\n",
    "                '{:1.2f}%'.format(0),\n",
    "                ha=\"center\",fontsize=15)\n",
    "    else:\n",
    "        total = sum((df.treatment==T_dict[i]))\n",
    "        height = p.get_height()\n",
    "        ax.text(p.get_x()+p.get_width()/2.,\n",
    "                height + 3,\n",
    "                '{:1.2f}%'.format(height/total*100),\n",
    "                ha=\"center\",fontsize=15) \n",
    "        \n",
    "        \n",
    "n_gave_T1 = sum(df[df.treatment=='T1'].contribution > 0.5)\n",
    "n_NOTgave_T1 = len(df[df.treatment=='T1']) - n_gave_T1\n",
    "n_gave_T2 = sum(df[df.treatment=='T2'].contribution > 0.5)\n",
    "n_NOTgave_T2 = len(df[df.treatment=='T2']) - n_gave_T2\n",
    "n_gave_T3 = sum(df[df.treatment=='T3'].contribution > 0.5)\n",
    "n_NOTgave_T3 = len(df[df.treatment=='T3']) - n_gave_T3\n",
    "n_gave_T4 = sum(df[df.treatment=='T4'].contribution == 0.5)\n",
    "n_NOTgave_T4 = len(df[df.treatment=='T4']) - n_gave_T4\n",
    "\n",
    "print('Between T1 and T2',proportions_ztest(n_gave_T1,n_NOTgave_T1+n_gave_T1,\n",
    "                                            value=n_gave_T2/(n_gave_T2+n_NOTgave_T2),\n",
    "                                            alternative='larger'))\n",
    "\n",
    "print('Between T1 and T3',proportions_ztest(n_gave_T1,n_NOTgave_T1+n_gave_T1,\n",
    "                                            value=n_gave_T3/(n_gave_T3+n_NOTgave_T3),\n",
    "                                            alternative='larger'))\n",
    "\n",
    "\n",
    "print('Between T2 and T3',proportions_ztest(n_gave_T2,n_NOTgave_T2+n_gave_T2,\n",
    "                                            value=n_gave_T3/(n_gave_T3+n_NOTgave_T3),\n",
    "                                            alternative='larger'))\n",
    "print('Between T3 and T4',proportions_ztest(n_gave_T3,n_NOTgave_T3+n_gave_T3,\n",
    "                                            value=n_gave_T4/(n_gave_T4+n_NOTgave_T4),\n",
    "                                            alternative='larger'))\n",
    "    \n",
    "ax.tick_params(labelsize=15)\n",
    "ax.set_xlabel('Condition',size=19)\n",
    "ax.set_ylabel('# Participants donated > %50',size=19)\n",
    "ax.set_ylim(0,370)\n",
    "#plt.savefig('./gave_more_50_percent.pdf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['ClickWorker', 'CrowdSource', 'CrowdFlower', 'microWorkers',\n",
       "       'Prolific', 'Other'], dtype='<U12')"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "platforms_occurances = []\n",
    "df['n_plaforms'] = 0\n",
    "for indx,participant in df.iterrows():\n",
    "    try:\n",
    "        platforms = list(filter(None, participant.otherPlatforms.split(',')))\n",
    "    except:\n",
    "        platforms = []\n",
    "    if 'Others' in platforms: platforms.remove('Others')\n",
    "    try:\n",
    "        additionalPlatforms = list(filter(None, participant.otherPlatforms_6_TEXT.split(',')))\n",
    "    except:\n",
    "        additionalPlatforms = []    \n",
    "    if (len(additionalPlatforms) + len(platforms)) == 0:\n",
    "        df.set_value(indx, 'n_plaforms', 1)\n",
    "    else:\n",
    "        df.set_value(indx, 'n_plaforms', len(additionalPlatforms) + len(platforms))\n",
    "    platforms_occurances += additionalPlatforms\n",
    "    platforms_occurances += platforms\n",
    "    \n",
    "platforms_occurances = list(filter(lambda a: a != 'Amazon Mechanical Turk', platforms_occurances))\n",
    "fig = plt.figure(figsize=(18, 6))\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "\n",
    "platforms_count = Counter(platforms_occurances)\n",
    "\n",
    "### change values to 'others' for the non common ones\n",
    "for index, item in enumerate(platforms_occurances):\n",
    "    if platforms_count[item] < 5:\n",
    "        platforms_occurances[index] = 'Other'\n",
    "platforms_count = Counter(platforms_occurances)\n",
    "platforms_order = np.array(platforms_count.most_common())[:,0]\n",
    "\n",
    "\n",
    "sns.countplot(x=platforms_occurances,ax=ax,\n",
    "              order=platforms_order)\n",
    "total = len(df)\n",
    "for i,p in enumerate(ax.patches):\n",
    "    height = p.get_height()\n",
    "    ax.text(p.get_x()+p.get_width()/2.,\n",
    "            height + 4,\n",
    "            '{:1.2f}%'.format(height/total*100),\n",
    "            ha=\"center\", fontsize=15) \n",
    "ax.tick_params(labelsize=15)\n",
    "#ax.set_xlabel('Platform',size=19)\n",
    "#plt.setp( ax.xaxis.get_majorticklabels(), rotation=15 )\n",
    "ax.set_ylabel('# Participants',size=19)\n",
    "ax.set_ylim(0,300)\n",
    "ax.set_xlabel('Membership in other microwork platforms',size=19)\n",
    "#sns.swarmplot(x=\"expecting_from_other\", y=\"contribution\", hue=\"treatment\", hue_order=order, data=df,ax=ax)\n",
    "\n",
    "plt.savefig('./other_platforms.pdf')\n",
    "# Use more informative axis labels than are provided by default\n",
    "#g.set_axis_labels(\"Sepal length (mm)\", \"Sepal width (mm)\")\n",
    "platforms_order"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, '# Participants')"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(5, 6))\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "#ax.axvline(df.age.median(),c='r')\n",
    "\n",
    "sns.countplot(x='n_plaforms', data=df,ax=ax,\n",
    "              order=range(1,6))\n",
    "total = len(df)\n",
    "for i,p in enumerate(ax.patches):\n",
    "    height = p.get_height()\n",
    "    ax.text(p.get_x()+p.get_width()/2.,\n",
    "            height + 10,\n",
    "            '{:1.2f}%'.format(height/total*100),\n",
    "            ha=\"center\", fontsize=15) \n",
    "ax.tick_params(labelsize=15)\n",
    "#ax.set_xlabel('Platform',size=19)\n",
    "ax.set_ylabel('# Participants',size=19)\n",
    "ax.set_ylim(0,1900)\n",
    "\n",
    "ax.tick_params(labelsize=15)\n",
    "ax.set_xlabel('Number of platforms',size=19)\n",
    "ax.set_ylabel('# Participants',size=19)\n",
    "#plt.savefig('./n_plaforms_distribution.pdf')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h1> Heterogeneous treatment effect (Moderators)</h1>\n",
    "<p>computed using the Causal tree approach by Susan Athey et al. in the R file</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>Experience on MTurk</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.0"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(df.workerDuration)\n",
    "df.workerDuration.median()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1501\n",
      "836\n"
     ]
    }
   ],
   "source": [
    "df['workExperience'] = np.nan\n",
    "df.loc[(df.workerDuration<=2), 'workExperience'] = 'new workers'\n",
    "df.loc[(df.workerDuration>2), 'workExperience'] = 'experienced workers'\n",
    "#df.loc[(df.workerDuration==5), 'workExperience'] = 'experienced workers'\n",
    "#df.loc[(df.workerDuration>=6), 'workExperience'] = 'experienced workers'\n",
    "#df['workExperience']\n",
    "print(sum(df['workExperience']=='new workers'))\n",
    "print(sum(df['workExperience']=='experienced workers'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(10, 6))\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "\n",
    "g = sns.barplot(x=\"workExperience\", y=\"donation_percentage\",hue='treatment',hue_order=order,\n",
    "                order=['new workers','experienced workers'], data=df,\n",
    "                ax=ax,ci=68, errwidth=1, capsize=0.1) \n",
    "\n",
    "#g = sns.barplot(x=\"treatment\", y=\"contribution\",hue='workExperience',hue_order=['new workers','experienced workers'],\n",
    "#                order=order, data=df,\n",
    "#                ax=ax,ci=68, errwidth=1, capsize=0.1) \n",
    "\n",
    "ax.tick_params(labelsize=15)\n",
    "ax.set_xlabel('Experience on MTurk',size=19)\n",
    "ax.set_ylabel('Avg. donation',size=19)\n",
    "ax.set_ylim(df.contribution.min(),20)\n",
    "plt.savefig('./avg_donation_experience_option1.pdf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "experienced workers\n",
      "average\n",
      "T1 0.12401662049861496\n",
      "T2 0.10901369863013706\n",
      "T3 0.09700000000000002\n",
      "T4 0.013211009174311925\n",
      "Median\n",
      "T1 0.0\n",
      "T2 0.0\n",
      "T3 0.0\n",
      "T4 0.0\n",
      "signiciance\n",
      "T1 > T2 p-val: MannwhitneyuResult(statistic=63625.0, pvalue=0.1766669007156213)\n",
      "T1 > T3 p-val: MannwhitneyuResult(statistic=66536.5, pvalue=0.013044060875718116)\n",
      "T2 > T3 p-val: MannwhitneyuResult(statistic=69638.5, pvalue=0.09297881209220249)\n",
      "T3 > T4 p-val: MannwhitneyuResult(statistic=16879.5, pvalue=1.2528494156899813e-06)\n",
      "% improvement (all not significant)\n",
      "T1 vs T2 0.13762418904233303\n",
      "T1 vs T3 0.2785218608104633\n",
      "T2 vs T3 0.12385256319728909\n"
     ]
    }
   ],
   "source": [
    "print('experienced workers')\n",
    "workerExperience = df.workExperience=='experienced workers'\n",
    "print('average')\n",
    "print('T1', df[T1_mask & workerExperience]['contribution'].mean())\n",
    "print('T2', df[T2_mask & workerExperience]['contribution'].mean())\n",
    "print('T3', df[T3_mask & workerExperience]['contribution'].mean())\n",
    "print('T4', df[T4_mask & workerExperience]['contribution'].mean())\n",
    "\n",
    "print('Median')\n",
    "print('T1', df[T1_mask & workerExperience]['contribution'].median())\n",
    "print('T2', df[T2_mask & workerExperience]['contribution'].median())\n",
    "print('T3', df[T3_mask & workerExperience]['contribution'].median())\n",
    "print('T4', df[T4_mask & workerExperience]['contribution'].median())\n",
    "\n",
    "print('signiciance')\n",
    "print('T1 > T2 p-val:',sp.mannwhitneyu(df[T1_mask & workerExperience]['contribution'],df[T2_mask & workerExperience]['contribution']))\n",
    "print('T1 > T3 p-val:',sp.mannwhitneyu(df[T1_mask & workerExperience]['contribution'],df[T3_mask & workerExperience]['contribution']))\n",
    "print('T2 > T3 p-val:',sp.mannwhitneyu(df[T2_mask & workerExperience]['contribution'],df[T3_mask & workerExperience]['contribution']))\n",
    "print('T3 > T4 p-val:',sp.mannwhitneyu(df[T3_mask & workerExperience]['contribution'],df[T4_mask & workerExperience]['contribution']))\n",
    "\n",
    "print('% improvement (all not significant)')\n",
    "print('T1 vs T2', (df[T1_mask & workerExperience]['contribution'].mean()-df[T2_mask & workerExperience]['contribution'].mean())/df[T2_mask & workerExperience]['contribution'].mean())\n",
    "print('T1 vs T3', (df[T1_mask & workerExperience]['contribution'].mean()-df[T3_mask & workerExperience]['contribution'].mean())/df[T3_mask & workerExperience]['contribution'].mean())\n",
    "print('T2 vs T3', (df[T2_mask & workerExperience]['contribution'].mean()-df[T3_mask & workerExperience]['contribution'].mean())/df[T3_mask & workerExperience]['contribution'].mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "less experienced workers\n",
      "average\n",
      "T1 0.17577342047930278\n",
      "T2 0.14886956521739128\n",
      "T3 0.11954545454545461\n",
      "T4 0.015211267605633804\n",
      "Median\n",
      "T1 0.01\n",
      "T2 0.0\n",
      "T3 0.0\n",
      "T4 0.0\n",
      "signiciance\n",
      "T1 > T2 p-val: MannwhitneyuResult(statistic=97560.0, pvalue=0.015225261942600534)\n",
      "T1 > T3 p-val: MannwhitneyuResult(statistic=85890.5, pvalue=9.147481765598285e-06)\n",
      "T2 > T3 p-val: MannwhitneyuResult(statistic=93629.0, pvalue=0.014469293359972683)\n",
      "T3 > T4 p-val: MannwhitneyuResult(statistic=22331.0, pvalue=2.1813441908874562e-10)\n",
      "T3 (exp vs non-exp): MannwhitneyuResult(statistic=56667.0, pvalue=0.17909776524503213)\n",
      "% improvement\n",
      "T1 vs T2 0.18072099036914852\n",
      "T1 vs T3 0.4703480040093761\n",
      "T2 vs T3 0.24529674326334847\n"
     ]
    }
   ],
   "source": [
    "print('less experienced workers')\n",
    "workerExperience = df.workExperience=='new workers'\n",
    "print('average')\n",
    "print('T1', df[T1_mask & workerExperience]['contribution'].mean())\n",
    "print('T2', df[T2_mask & workerExperience]['contribution'].mean())\n",
    "print('T3', df[T3_mask & workerExperience]['contribution'].mean())\n",
    "print('T4', df[T4_mask & workerExperience]['contribution'].mean())\n",
    "\n",
    "print('Median')\n",
    "print('T1', df[T1_mask & workerExperience]['contribution'].median())\n",
    "print('T2', df[T2_mask & workerExperience]['contribution'].median())\n",
    "print('T3', df[T3_mask & workerExperience]['contribution'].median())\n",
    "print('T4', df[T4_mask & workerExperience]['contribution'].median())\n",
    "\n",
    "print('signiciance')\n",
    "print('T1 > T2 p-val:',sp.mannwhitneyu(df[T1_mask & workerExperience]['contribution'],df[T2_mask & workerExperience]['contribution']))\n",
    "print('T1 > T3 p-val:',sp.mannwhitneyu(df[T1_mask & workerExperience]['contribution'],df[T3_mask & workerExperience]['contribution']))\n",
    "print('T2 > T3 p-val:',sp.mannwhitneyu(df[T2_mask & workerExperience]['contribution'],df[T3_mask & workerExperience]['contribution']))\n",
    "print('T3 > T4 p-val:',sp.mannwhitneyu(df[T3_mask & workerExperience]['contribution'],df[T4_mask & workerExperience]['contribution']))\n",
    "print('T3 (exp vs non-exp):',sp.mannwhitneyu(df[T3_mask & workerExperience]['contribution'],df[T3_mask & ~workerExperience]['contribution']))\n",
    "\n",
    "\n",
    "print('% improvement')\n",
    "print('T1 vs T2', (df[T1_mask & workerExperience]['contribution'].mean()-df[T2_mask & workerExperience]['contribution'].mean())/df[T2_mask & workerExperience]['contribution'].mean())\n",
    "print('T1 vs T3', (df[T1_mask & workerExperience]['contribution'].mean()-df[T3_mask & workerExperience]['contribution'].mean())/df[T3_mask & workerExperience]['contribution'].mean())\n",
    "print('T2 vs T3', (df[T2_mask & workerExperience]['contribution'].mean()-df[T3_mask & workerExperience]['contribution'].mean())/df[T3_mask & workerExperience]['contribution'].mean())\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Appendix t-test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "average\n",
      "T1 0.15592274678111584\n",
      "T2 0.13572230014025236\n",
      "T3 0.11582743988684586\n",
      "T4 0.013394495412844038\n",
      "Median\n",
      "T1 0.0\n",
      "T2 0.0\n",
      "T3 0.0\n",
      "T4 0.0\n",
      "signiciance\n",
      "T1 > T2 p-val: 0.898207180732\n",
      "T1 > T3 p-val: 1.85144259901\n",
      "T2 > T3 p-val: 0.922786561603\n",
      "T3 > T4 p-val: 3.84853764601\n"
     ]
    }
   ],
   "source": [
    "print('average')\n",
    "print('T1', df[T1_mask]['contribution'].mean())\n",
    "print('T2', df[T2_mask]['contribution'].mean())\n",
    "print('T3', df[T3_mask]['contribution'].mean())\n",
    "print('T4', df[T4_mask]['contribution'].mean())\n",
    "\n",
    "print('Median')\n",
    "print('T1', df[T1_mask]['contribution'].median())\n",
    "print('T2', df[T2_mask]['contribution'].median())\n",
    "print('T3', df[T3_mask]['contribution'].median())\n",
    "print('T4', df[T4_mask]['contribution'].median())\n",
    "\n",
    "#divide by 2 because the one-sided tests can be backed out from the two-sided tests. (With symmetric distributions one-sided p-value is just half of the two-sided pvalue)\n",
    "print('signiciance')\n",
    "print('T1 > T2 p-val:',sp.ttest_ind(df[T1_mask]['contribution'],df[T2_mask]['contribution'])[0]/2.0)\n",
    "print('T1 > T3 p-val:',sp.ttest_ind(df[T1_mask]['contribution'],df[T3_mask]['contribution'])[0]/2.0)\n",
    "print('T2 > T3 p-val:',sp.ttest_ind(df[T2_mask]['contribution'],df[T3_mask]['contribution'])[0]/2.0)\n",
    "print('T3 > T4 p-val:',sp.ttest_ind(df[T3_mask]['contribution'],df[T4_mask]['contribution'])[0]/2.0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>sensitivity power analyses</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from statsmodels.stats.power import tt_ind_solve_power"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "T1 > T2\n",
      "minimum detectable standrized effect: 0.07033720374176824\n",
      "actual standrized effect T1 > T2: 0.09554281694264043\n"
     ]
    }
   ],
   "source": [
    "# The standardized effect size ie. difference between the two means divided by the standard deviation\n",
    "#mean_diff = df[T1_mask]['donation_percentage'].mean() /  df[T1_mask]['donation_percentage'].std() - df[T2_mask]['donation_percentage'].mean()/df[T2_mask]['donation_percentage'].std()\n",
    "\n",
    "##or?\n",
    "mean_diff = (df[T1_mask]['donation_percentage'].mean() - df[T2_mask]['donation_percentage'].mean())/df[T1_mask | T2_mask]['donation_percentage'].std()\n",
    "\n",
    "\n",
    "min_detect_effect = tt_ind_solve_power(nobs1=len(T1_mask), alpha=0.05, power=0.8, \n",
    "               ratio=len(T1_mask)/len(T2_mask), alternative='larger')\n",
    "\n",
    "print(\"T1 > T2\")\n",
    "print(\"minimum detectable standrized effect:\", min_detect_effect)\n",
    "print(\"actual standrized effect T1 > T2:\", mean_diff)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "T1 > T3\n",
      "minimum detectable standrized effect: 0.07033720374176824\n",
      "actual standrized effect T1 > T2: 0.19662053155641868\n"
     ]
    }
   ],
   "source": [
    "# The standardized effect size ie. difference between the two means divided by the standard deviation\n",
    "#mean_diff = df[T1_mask]['donation_percentage'].mean() /  df[T1_mask]['donation_percentage'].std() - df[T3_mask]['donation_percentage'].mean()/df[T3_mask]['donation_percentage'].std()\n",
    "\n",
    "##or?\n",
    "mean_diff = (df[T1_mask]['donation_percentage'].mean() - df[T3_mask]['donation_percentage'].mean())/df[T1_mask | T3_mask]['donation_percentage'].std()\n",
    "\n",
    "\n",
    "min_detect_effect = tt_ind_solve_power(nobs1=len(T1_mask), alpha=0.05, power=0.8, \n",
    "               ratio=len(T1_mask)/len(T3_mask), alternative='larger')\n",
    "\n",
    "print(\"T1 > T3\")\n",
    "print(\"minimum detectable standrized effect:\", min_detect_effect)\n",
    "print(\"actual standrized effect T1 > T2:\", mean_diff)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "T2 > T3\n",
      "minimum detectable standrized effect: 0.07033720374176824\n",
      "actual standrized effect T1 > T2: 0.09787078270062913\n"
     ]
    }
   ],
   "source": [
    "# The standardized effect size ie. difference between the two means divided by the standard deviation\n",
    "#mean_diff = df[T2_mask]['donation_percentage'].mean() /  df[T2_mask]['donation_percentage'].std() - df[T3_mask]['donation_percentage'].mean()/df[T3_mask]['donation_percentage'].std()\n",
    "\n",
    "##or?\n",
    "mean_diff = (df[T2_mask]['donation_percentage'].mean() - df[T3_mask]['donation_percentage'].mean())/df[T2_mask | T3_mask]['donation_percentage'].std()\n",
    "\n",
    "\n",
    "min_detect_effect = tt_ind_solve_power(nobs1=len(T2_mask), alpha=0.05, power=0.8, \n",
    "               ratio=len(T2_mask)/len(T3_mask), alternative='larger')\n",
    "\n",
    "print(\"T2 > T3\")\n",
    "print(\"minimum detectable standrized effect:\", min_detect_effect)\n",
    "print(\"actual standrized effect T1 > T2:\", mean_diff)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "T3 > T4\n",
      "minimum detectable standrized effect: 0.07033720374176824\n",
      "actual standrized effect T1 > T2: 0.5783419923381494\n"
     ]
    }
   ],
   "source": [
    "# The standardized effect size ie. difference between the two means divided by the standard deviation\n",
    "#mean_diff = df[T2_mask]['donation_percentage'].mean() /  df[T2_mask]['donation_percentage'].std() - df[T3_mask]['donation_percentage'].mean()/df[T3_mask]['donation_percentage'].std()\n",
    "\n",
    "##or?\n",
    "mean_diff = (df[T3_mask]['donation_percentage'].mean() - df[T4_mask]['donation_percentage'].mean())/df[T3_mask | T4_mask]['donation_percentage'].std()\n",
    "\n",
    "\n",
    "min_detect_effect = tt_ind_solve_power(nobs1=len(T3_mask), alpha=0.05, power=0.8, \n",
    "               ratio=len(T3_mask)/len(T4_mask), alternative='larger')\n",
    "\n",
    "print(\"T3 > T4\")\n",
    "print(\"minimum detectable standrized effect:\", min_detect_effect)\n",
    "print(\"actual standrized effect T1 > T2:\", mean_diff)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}
